CN109802394A - A kind of probability load flow calculation method counted and distributed generation resource and electric car access - Google Patents
A kind of probability load flow calculation method counted and distributed generation resource and electric car access Download PDFInfo
- Publication number
- CN109802394A CN109802394A CN201910255111.XA CN201910255111A CN109802394A CN 109802394 A CN109802394 A CN 109802394A CN 201910255111 A CN201910255111 A CN 201910255111A CN 109802394 A CN109802394 A CN 109802394A
- Authority
- CN
- China
- Prior art keywords
- node
- formula
- electric car
- power
- probability
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The probability load flow calculation method of of the invention a kind of meter and distributed generation resource and electric car access, comprising: establish the probabilistic model of distributed electrical source node Yu electric car node；According to traditional Load flow calculation, the electrical network basic datas such as power grid connection topology, line impedance, node injecting power are obtained；Input variable is sampled to obtain sample matrix using the Latin Hypercube Sampling method based on Random Walk；Each rank cumulant of input variable is calculated according to sample matrix and cumulant and the relationship of origin matrix；The cumulative distribution function of state variable and Branch Power Flow is acquired by series expansion；Carry out Cumulants method probabilistic load flow.The present invention not only allows for the uncertain power injection of photovoltaic in power distribution network, wind-powered electricity generation, it is also considered that the random load by taking electric car as an example makes the new energy receiving ability that the more accurate computational accuracy of result is high, sample rate is fast, facilitates promotion power train.
Description
Technical field
The invention belongs to probability load flow calculation method field, it is related to a kind of meter and distributed generation resource and electric car access
Probability load flow calculation method.
Background technique
China is world energy consumption third big country, and with the rapid development of productivity, energy-output ratio is also constantly increasing
Add, wherein the consumption of petroleum, coal occupies enormous proportions.But as global climate gradual warms up in recent years, the day of petroleum resources
Beneficial exhausted, the exploitation of new energy seems extremely urgent.From domestic wind-powered electricity generation, photovoltaic respectively at implementation mark post electricity in 2009,2013
Since valence system (Benchmark electricity price), domestic New Energy Industry rapid development.National photovoltaic in 2017
Generator installation reaches 130000MW, and national wind-power electricity generation installation reaches 130780MW.
There are a large amount of uncertain factors, such as load fluctuation, system equipment failure in practical power systems operational process.
The development of the random loads such as large-scale distributed power grid and electric car further increases the uncertainty of system.Probability tide
Stream calculation can analyze electric network swim characteristic, be that power system stability is run using various uncertain factors as input variable
Important tool.
Probability load flow calculation method traditional at present has Monte Carlo Method, point estimations and convolution method.It wherein covers special
Calot's method generates the sample with correlation with sampling techniques, then carries out multiple certainty Load flow calculation, obtains output variable
Statistical distribution characteristic, but the method calculation amount is complex, needs to be sampled on a large scale, elapsed time cost.Point
The estimation technique is to seek the probabilistic method of output each rank square of stochastic variable according to the probability distribution of known input stochastic variable,
But the method is unable to get the cumulative distribution function of output variable and exports the order of accuarcy of each rank square of stochastic variable with order
Increase and reduce.Convolution method is to carry out convolutional calculation according to the probability-distribution function of input stochastic variable, and it is random to acquire output
The probability density characteristics of variable, but the method needs to be calculated with Convolution Formula, and mathematical computations are complicated.
Summary of the invention
In order to solve the above technical problems, the present invention provides a kind of probability tide counted and distributed generation resource and electric car access
Flow calculation methodologies, computational accuracy is high, sample rate is fast, helps to promote the new energy receiving ability of power train.
The present invention provides a kind of probability load flow calculation method counted and distributed generation resource and electric car access, including as follows
Step:
Step 1: the probabilistic model of distributed wind-powered electricity generation node, distributed photovoltaic node and electric car node is established, with
Basic electric network model collectively forms system model；
Step 2: the basic data of system model, including power grid connection topology, route resistance are obtained according to traditional Load flow calculation
Anti-, load injecting power, engine injecting power, the stochastic variable of each distributed electrical source node and electric car node it is random
Variable；
Step 3: setting system model interior joint type ignores distributed wind-powered electricity generation node, distributed photovoltaic node and electricity
Electrical automobile node obtains normal condition variable with traditional Newton-Laphson method to system being determined property Load flow calculation；
Step 4: probabilistic model being sampled to obtain phase using the Latin Hypercube Sampling method based on Random Walk
Answer the sample matrix of probabilistic model；
Step 5: seeking the sensitivity matrix of sample matrix；
Step 6: according to the relationship of sample matrix and moment of the orign, each rank of probabilistic model being calculated partly not based on Cumulants method
Variable；
Step 7: according to the sensitivity matrix of basic data, normal condition variable, sample matrix and sample matrix, using
Cumulants method carries out probabilistic load flow to system model.
In the probability load flow calculation method of meter and distributed generation resource and electric car access of the invention, the step 1
The middle probabilistic model for establishing distributed wind-powered electricity generation node specifically:
Mean wind speed obeys Weibull distribution:
In formula: v is mean wind speed, and k is form parameter, and c is dimensional parameters；
The active power output characteristics of Wind turbines is as follows:
In formula: vciTo cut wind speed；vrateFor rated wind speed；vcoFor cut-out wind speed；PrateFor Wind turbines nominal output；
By formula (1) (2) it can be concluded that the probability distribution of output of wind electric field:
In formula,
In the probability load flow calculation method of meter and distributed generation resource and electric car access of the invention, the step 1
The middle probabilistic model for establishing distributed photovoltaic node specifically:
Intensity of illumination obeys Beta distribution, pdf model are as follows:
In formula: r is actual intensity of illumination；rmaxFor intensity of illumination maximum in search time；α and β is the shape of Beta distribution
Shape parameter；Γ () is Gamma function；
Equipped with n block photovoltaic panel, every piece of photovoltaic plate suqare is Ai, transfer efficiency ηi, the active power of photovoltaic panel is expressed as:
According to the probability density for the active power that formula (4), (5), (6) available photovoltaic export are as follows:
Pmax=A η rmax (8)
In formula: PmaxFor the maximum output of photovoltaic plant active power.
In the probability load flow calculation method of meter and distributed generation resource and electric car access of the invention, the step 1
The middle probabilistic model for establishing electric car node specifically:
Electric car model is established, the electric car probabilistic model used is Poisson distribution:
In formula: λEV,tThe electric car quantity desired value accessed by t moment, nEV,tThe electric car accessed by t moment
Quantity.
In the probability load flow calculation method of meter and distributed generation resource and electric car access of the invention, the step 4
Specifically:
Step 4.1: determining the stochastic variable number of distribution wind-powered electricity generation node, distributed photovoltaic node and electric car node
Measure K；
Step 4.2: determining the sampling number N of Latin Hypercube Sampling method；
Step 4.3: each probabilistic model being sampled using Latin Hypercube Sampling method, generates the original sample of K × N
Matrix S 'ij, wherein i represents i-th of stochastic variable, and j represents stochastic variable jth time sampled value；
Step 4.4: using the method for walk random to original sample matrix S 'ijIt resequences, obtains final sample
Matrix Sij。
In the probability load flow calculation method of meter and distributed generation resource and electric car access of the invention, the step 5
Specifically:
Step 5.1: Load flow calculation equation uses following polar form:
In formula (12): W is node injecting power；H is Branch Power Flow；X is node voltage value；
Step 5.2: formula (12) is rewritten as to the linearisation AC power flow equation of matrix expression:
In formula (13): X0、W0、H0Respectively indicate the desired value of node voltage, the desired value of node injecting power and branch tide
The desired value of stream, and meet:
△ X is the variable quantity of node voltage；△ W is the random perturbation of node injecting power；△ H is the variation of Branch Power Flow
Amount；
Step 5.3: the high-order term for carrying out Taylor expansion, and ignore 2 ranks or more formula (13) obtains:
Step 5.4: formula (15) is converted and can be obtained:
In formula: J0For the Jacobian matrix of basic electric network model；S0For J0Inverse matrix；G0For 2b × 2n rank matrix, wherein
B is the circuitry number of system model, and n is the number of nodes of system model；
Step 5.5: in the case where system operates normally, with traditional Newton-Raphson Load flow calculation, being existed by calculating
The desired value X of the node voltage of datum mark operation0, Branch Power Flow desired value H0With Jacobian matrix J0, further seek sensitive
Spend matrix T0。
In the probability load flow calculation method of meter and distributed generation resource and electric car access of the invention, the step 6
Specifically:
Step 6.1: the final sample matrix S obtained according to Latin Hypercube Sampling methodijIn N number of sample [X1,
X2,···,XN], calculate separately each rank moment of the orign α of each samplev:
Step 6.2: further according to the relationship of cumulant and moment of the orign, finding out each rank cumulant γ of N number of samplev:
Take preceding 7 rank cumulant to guarantee higher computational accuracy.
In the probability load flow calculation method of meter and distributed generation resource and electric car access of the invention, the step 7
Specifically:
Step 7.1: using half step quantity method Load flow calculation, the stochastic variable Δ W of the injecting power of each node mmIt can be with table
It is shown as:
In formula:Represent convolution algorithm；ΔWGmWith Δ WLmRespectively represent the generator power and load power of node m
Stochastic variable；
Step 7.2: according to the homogeneity and additive property of Cumulants method, formula (20) being rewritten as to the algebra of Cumulants method
Operation, the k rank cumulant of node m injecting powerIt can indicate are as follows:
Step 7.3: according to the linearisation tide model of formula (16), showing that the k rank half of node voltage and Branch Power Flow is constant
Measure Δ X(k)With Δ H(k):
Step 7.4: solving the probability of the cumulant of stochastic variable respectively with the method for Gram-Clarlier series expansion
The probability density of the cumulant of density, the probability density of the cumulant of node voltage and Branch Power Flow:
Cumulative distribution function f (x) can be expressed as the series expansion of the all-order derivative containing α (x), then each rank of f (x) is led
Number expansion are as follows:
In formula, x is each stochastic variable, and α (x) is the probability density function of the stochastic variable of standardized normal distribution；
H in formular(x) it is Xue Fu-Amire spy's multinomial:
Convolution (23) (24) obtains the cumulative distribution function of stochastic variable are as follows:
G is input variable, node voltage and Branch Power Flow and cumulant in formula.
Step 7.5: according to cumulative distribution function and probability density relationship:
F (x) dx=P (26)
Find out the probability distribution of node voltage, power；It is embodied as a certain node a certain moment each voltage, performance number
Probability and all moment voltages of a certain node day, power probability maximum value.
The probability load flow calculation method of of the invention a kind of meter and distributed generation resource and electric car access, at least have with
It is lower the utility model has the advantages that
Stochastic variable is carried out using the Latin Hypercube Sampling method based on Random Walk in calculating process of the present invention
Sampling, Random Walk and traditional Latin Hypercube Sampling method are combined, the lower sampling matrix of correlation is obtained,
And Cholesky decomposition is carried out to sampling matrix, the correlation between stochastic variable is further eliminated, avoids having used Meng Teka
Lip river method carries out correlation processing to node power injection rate, ensure that sampling computational accuracy and improves sample rate.
Cumulant probability load flow calculation method is used in calculating process of the present invention, replaces convolutional calculation with additional calculation,
Simplify algorithm.Wherein power flow equation is current equation, compared with the power equation used in conventional probabilistic load flow, electric current
Equation is that linear equation is more in line with Cumulants method probabilistic load flow.
The present invention calculates the uncertain power injection for not only allowing for photovoltaic in power distribution network, wind-powered electricity generation, it is also considered that with electronic
Random load for automobile keeps result more accurate.
The present invention, which calculates, can be suitable for complex electric network system of the large-scale distributed power supply with electric car after grid-connected
Operating analysis, safe early warning etc. are of great significance, and the new energy for helping to be promoted power train receives ability.
Detailed description of the invention
Fig. 1 be it is a kind of meter and distributed generation resource and electric car access probability load flow calculation method flow chart.
Specific embodiment
For overcome the deficiencies in the prior art, the present invention provide it is a kind of meter and distributed generation resource accessed with electric car it is general
Rate tidal current computing method.Consider the Latin Hypercube Sampling method (Latin for being based on Random Walk (Random Walk, RW)
Hypercube sampling, LHS), initially set up the probabilistic model of wind-powered electricity generation, photovoltaic distributed generation resource and electric car；Secondly
Random Walk is obtained into sample matrix with the method that traditional Latin Hypercube Sampling method combines, eliminate stochastic variable it
Between correlation, improve calculating speed while guaranteeing computational accuracy；Finally Probabilistic Load Flow meter is carried out using Cumulants method
It calculates.By replacing convolution algorithm using algebraic operation in the case where input variable is mutually indepedent, operation is simple, speed is fast,
It being capable of quick and precisely assessment system operation characteristic.
The probability load flow calculation method of of the invention a kind of meter and distributed generation resource and electric car access as shown in Figure 1,
Include the following steps:
Step 1: the probabilistic model of distributed wind-powered electricity generation node, distributed photovoltaic node and electric car node is established, with
Basic electric network model collectively forms system model；
(1) probabilistic model of distributed wind-powered electricity generation node is established specifically:
Mean wind speed obeys Weibull distribution:
In formula: v is mean wind speed, and k is form parameter, and c is dimensional parameters；
The active power output characteristics of Wind turbines is as follows:
In formula: vciTo cut wind speed；vrateFor rated wind speed；vcoFor cut-out wind speed；PrateFor Wind turbines nominal output；
By formula (1) (2) it can be concluded that the probability distribution of output of wind electric field:
In formula,
(2) probabilistic model of distributed photovoltaic node is established specifically:
Intensity of illumination obeys Beta distribution, pdf model are as follows:
In formula: r is actual intensity of illumination；rmaxFor intensity of illumination maximum in search time；α and β is the shape of Beta distribution
Shape parameter；Γ () is Gamma function；
Equipped with n block photovoltaic panel, every piece of photovoltaic plate suqare is Ai, transfer efficiency ηi, the active power of photovoltaic panel is expressed as:
According to the probability density for the active power that formula (4), (5), (6) available photovoltaic export are as follows:
Pmax=A η rmax (8)
In formula: PmaxFor the maximum output of photovoltaic plant active power.
(3) probabilistic model of electric car node is established specifically:
Electric car model is established, the electric car probabilistic model that the invention patent uses is approximately Poisson distribution:
In formula: λEV,tThe electric car quantity desired value accessed by t moment, nEV,tThe electric car accessed by t moment
Quantity.
Step 2: the basic data of system model, including power grid connection topology, route resistance are obtained according to traditional Load flow calculation
Anti-, load injecting power, engine injecting power, the stochastic variable of each distributed electrical source node and electric car node it is random
Variable；
Step 3: setting system model interior joint type ignores distributed wind-powered electricity generation node, distributed photovoltaic node and electricity
Electrical automobile node obtains normal condition variable with traditional Newton-Laphson method to system being determined property Load flow calculation；
When it is implemented, setting system model interior joint type, such as 1,2,3, k node be PV node, k+1,
K+2, n-1 node are PQ node, and n is balance nodes.Ignore distributed generation resource and random load node, with traditional
Newton-Laphson method obtains the normal condition variable of system to system being determined property Load flow calculation, and makees following label: V table
Show that node voltage, I indicate that node current, P indicate the active power of node injection, Q indicates the reactive power of node injection, △ table
Show amount of state variation, subscript G indicates PV node, and subscript L indicates that PQ node, subscript r indicate real part, and subscript m indicates imaginary part, Y table
Show admittance matrix.
Step 4: probabilistic model being sampled to obtain phase using the Latin Hypercube Sampling method based on Random Walk
Answer the sample matrix of probabilistic model, the step 4 specifically:
Step 4.1: determining the stochastic variable number of distribution wind-powered electricity generation node, distributed photovoltaic node and electric car node
Measure K；
Step 4.2: determining the sampling number N of Latin Hypercube Sampling method；
Step 4.3: each probabilistic model being sampled using Latin Hypercube Sampling method, generates the original sample of K × N
Matrix S 'ij, wherein i represents i-th of stochastic variable, and j represents stochastic variable jth time sampled value；
Step 4.4: using the method for walk random to original sample matrix S 'ijIt resequences, obtains final sample
Matrix Sij。
When it is implemented, to original sample matrix S 'ijIt resequences, the method that the present invention uses walk random should
Method is a kind of equiprobable algorithm, using stochastic variable as research object, carries out walk random in a matrix, because of its each side
To be all it is equiprobable, the method for walk random is applied in Latin hypercube sequence step, each stochastic variable is reduced
Correlation, it can be ensured that the randomness of each sample, isotropism and without Preference.
(1) Latin hypercube original sample matrix S ' is setijPrimary iteration point S0, the step size mu of walking.
(2) in order to generate suitable iteration control intersection number r, setting recycles initial value a=1.
(3) as a < r, new matrix L=[L of j random sequence is generated at random1,L2,…Li,Lj], and from this j
By following objective function in a new sequence, the smallest sequence S of the degree of correlation is selected1, complete a wherein walking and walk.
(4) specific objective function is as follows:
Cov () is covariance in formula；Var () is variance, L1,L2,…Li,LjIndicate each random sequence.
Formula (11) is to reduce sampled value L between each stochastic variablei,LjCorrelation.
The functional value of calculating is brought into (10) (11), if calculated value is more preferable than initial value, selects adopting for sample matrix
Sample value is put into sample matrix.Otherwise a=a+1 returns to step (3).
(5) if continuous r times all can not find optimal value, step size mu is halved into return step 1, the sequence of a weight new round at this time.Again
The sampled value that above-mentioned steps generate is brought into objective function.
The sampling matrix S ultimately producedij, wherein i represents i-th of stochastic variable, and j represents stochastic variable jth time sampling
Value.
Step 5: seeking the sensitivity matrix of sample matrix；
The present invention is using linearisation power flow equation, including node injecting power equation and Branch Power Flow equation, using exchange
Inearized model, by the high-order term that it carries out Taylor expansion, and ignore 2 ranks or more in benchmark operating point, specifically:
Step 5.1: Load flow calculation equation uses following polar form:
In formula (12): W is node injecting power；H is Branch Power Flow；X is node voltage value；
Step 5.2: formula (12) is rewritten as to the linearisation AC power flow equation of matrix expression:
In formula (13): X0、W0、H0Respectively indicate the desired value of node voltage, the desired value of node injecting power and branch tide
The desired value of stream, and meet:
△ X is the variable quantity of node voltage；△ W is the random perturbation of node injecting power；△ H is the variation of Branch Power Flow
Amount；
Step 5.3: the high-order term for carrying out Taylor expansion, and ignore 2 ranks or more formula (13) obtains:
Step 5.4: formula (15) is converted and can be obtained:
In formula: J0For the Jacobian matrix of basic electric network model；S0For J0Inverse matrix；G0For 2b × 2n rank matrix, wherein
B is the circuitry number of system model, and n is the number of nodes of system model；
Step 5.5: in the case where system operates normally, with traditional Newton-Raphson Load flow calculation, being existed by calculating
The desired value X of the node voltage of datum mark operation0, Branch Power Flow desired value H0With Jacobian matrix J0, further seek sensitive
Spend matrix T0。
Step 6: according to the relationship of sample matrix and moment of the orign, each rank of probabilistic model being calculated partly not based on Cumulants method
Variable, the step 6 specifically:
Step 6.1: the final sample matrix S obtained according to Latin Hypercube Sampling methodijIn N number of sample [X1,
X2,···,XN], calculate separately each rank moment of the orign α of each samplev:
Step 6.2: further according to the relationship of cumulant and moment of the orign, finding out each rank cumulant γ of N number of samplev:
Take preceding 7 rank cumulant to guarantee higher computational accuracy.
Step 7: according to the sensitivity matrix of basic data, normal condition variable, sample matrix and sample matrix, using
Cumulants method carries out probabilistic load flow, the step 7 to system model specifically:
Step 7.1: using half step quantity method Load flow calculation, the stochastic variable Δ W of the injecting power of each node mmIt can be with table
It is shown as:
In formula:Represent convolution algorithm；ΔWGmWith Δ WLmRespectively represent the generator power and load power of node m
Stochastic variable；
Step 7.2: according to the homogeneity and additive property of Cumulants method, formula (20) being rewritten as to the algebra of Cumulants method
Operation, the k rank cumulant of node m injecting powerIt can indicate are as follows:
Step 7.3: according to the linearisation tide model of formula (16), by showing that k rank linearizes tide model:
The independence of Cumulants method by the property of convolution on condition that determined, because using improved Latin in the present invention
The sampling matrix that hypercube obtains has eliminated stochastic variable correlation, so not needing to carry out stochastic variable further
Processing；
Step 7.4: solving the probability of the cumulant of stochastic variable respectively with the method for Gram-Clarlier series expansion
The probability density of the cumulant of density, the probability density of the cumulant of node voltage and Branch Power Flow:
Cumulative distribution function f (x) can be expressed as the series expansion of the all-order derivative containing α (x), then each rank of f (x) is led
Number expansion are as follows:
In formula, x is each stochastic variable, and α (x) is the probability density function of the stochastic variable of standardized normal distribution；
H in formular(x) it is Xue Fu-Amire spy's multinomial:
Convolution (23) (24) obtains the cumulative distribution function of stochastic variable are as follows:
G is input variable, node voltage and Branch Power Flow and cumulant in formula.
Step 7.5: according to cumulative distribution function and probability density relationship:
F (x) dx=P (26)
Find out the probability distribution of node voltage, power；It is embodied as a certain node a certain moment each voltage, performance number
Probability and all moment voltages of a certain node day, power probability maximum value.
The foregoing is merely presently preferred embodiments of the present invention, the thought being not intended to limit the invention, all of the invention
Within spirit and principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (8)
1. the probability load flow calculation method of a kind of meter and distributed generation resource and electric car access, which is characterized in that including as follows
Step:
Step 1: the probabilistic model of distributed wind-powered electricity generation node, distributed photovoltaic node and electric car node is established, with basis
Electric network model collectively forms system model；
Step 2: the basic data of system model is obtained according to traditional Load flow calculation, including power grid connects topology, line impedance, bears
Lotus injecting power, engine injecting power, the stochastic variable of each distributed electrical source node and electric car node stochastic variable；
Step 3: setting system model interior joint type ignores distributed wind-powered electricity generation node, distributed photovoltaic node and electronic vapour
Vehicle node obtains normal condition variable with traditional Newton-Laphson method to system being determined property Load flow calculation；
Step 4: probabilistic model being sampled to obtain using the Latin Hypercube Sampling method based on Random Walk corresponding general
The sample matrix of rate model；
Step 5: seeking the sensitivity matrix of sample matrix；
Step 6: according to the relationship of sample matrix and moment of the orign, each rank half for calculating probabilistic model based on Cumulants method is constant
Amount；
Step 7: according to the sensitivity matrix of basic data, normal condition variable, sample matrix and sample matrix, using partly not
Quantity method carries out probabilistic load flow to system model.
2. the probability load flow calculation method of meter as described in claim 1 and distributed generation resource and electric car access, feature
It is, the probabilistic model of distributed wind-powered electricity generation node is established in the step 1 specifically:
Mean wind speed obeys Weibull distribution:
In formula: v is mean wind speed, and k is form parameter, and c is dimensional parameters；
The active power output characteristics of Wind turbines is as follows:
In formula: vciTo cut wind speed；vrateFor rated wind speed；vcoFor cut-out wind speed；PrateFor Wind turbines nominal output；
By formula (1) (2) it can be concluded that the probability distribution of output of wind electric field:
In formula,
3. the probability load flow calculation method of meter as described in claim 1 and distributed generation resource and electric car access, feature
It is, the probabilistic model of distributed photovoltaic node is established in the step 1 specifically:
Intensity of illumination obeys Beta distribution, pdf model are as follows:
In formula: r is actual intensity of illumination；rmaxFor intensity of illumination maximum in search time；α and β is the shape ginseng of Beta distribution
Number；Γ () is Gamma function；
Equipped with n block photovoltaic panel, every piece of photovoltaic plate suqare is Ai, transfer efficiency ηi, the active power of photovoltaic panel is expressed as:
According to the probability density for the active power that formula (4), (5), (6) available photovoltaic export are as follows:
Pmax=A η rmax (8)
In formula: PmaxFor the maximum output of photovoltaic plant active power.
4. the probability load flow calculation method of meter as described in claim 1 and distributed generation resource and electric car access, feature
It is, the probabilistic model of electric car node is established in the step 1 specifically:
Electric car model is established, the electric car probabilistic model used is Poisson distribution:
In formula: λEV, tThe electric car quantity desired value accessed by t moment, nEV, tThe electric car number accessed by t moment
Amount.
5. the probability load flow calculation method of meter as described in claim 1 and distributed generation resource and electric car access, feature
It is, the step 4 specifically:
Step 4.1: determining the stochastic variable quantity K of distribution wind-powered electricity generation node, distributed photovoltaic node and electric car node；
Step 4.2: determining the sampling number N of Latin Hypercube Sampling method；
Step 4.3: each probabilistic model being sampled using Latin Hypercube Sampling method, generates the original sample matrix of K × N
S’ij, wherein i represents i-th of stochastic variable, and j represents stochastic variable jth time sampled value；
Step 4.4: using the method for walk random to original sample matrix S 'ijIt resequences, obtains final sample matrix
Sij。
6. the probability load flow calculation method of meter as described in claim 1 and distributed generation resource and electric car access, feature
It is, the step 5 specifically:
Step 5.1: Load flow calculation equation uses following polar form:
In formula (12): W is node injecting power；H is Branch Power Flow；X is node voltage value；
Step 5.2: formula (12) is rewritten as to the linearisation AC power flow equation of matrix expression:
In formula (13): X0、W0、H0Respectively indicate the desired value of node voltage, the desired value of node injecting power and Branch Power Flow
Desired value, and meet:
Δ X is the variable quantity of node voltage；Δ W is the random perturbation of node injecting power；Δ H is the variable quantity of Branch Power Flow；
Step 5.3: the high-order term for carrying out Taylor expansion, and ignore 2 ranks or more formula (13) obtains:
Step 5.4: formula (15) is converted and can be obtained:
In formula: J0For the Jacobian matrix of basic electric network model；S0For J0Inverse matrix；G0For 2b × 2n rank matrix, wherein b is to be
The circuitry number of system model, n are the number of nodes of system model；
Step 5.5: in the case where system operates normally, with traditional Newton-Raphson Load flow calculation, by calculating in benchmark
The desired value X of the node voltage of point operation0, Branch Power Flow desired value H0With Jacobian matrix J0, further seek sensitivity square
Battle array T0。
7. the probability load flow calculation method of meter as claimed in claim 5 and distributed generation resource and electric car access, feature
It is, the step 6 specifically:
Step 6.1: the final sample matrix S obtained according to Latin Hypercube Sampling methodijIn N number of sample [X1, X2..., XN],
Calculate separately each rank moment of the orign α of each samplev:
Step 6.2: further according to the relationship of cumulant and moment of the orign, finding out each rank cumulant γ of N number of samplev:
Take preceding 7 rank cumulant to guarantee higher computational accuracy.
8. the probability load flow calculation method of meter as claimed in claim 6 and distributed generation resource and electric car access, feature
It is, the step 7 specifically:
Step 7.1: using half step quantity method Load flow calculation, the stochastic variable Δ W of the injecting power of each node mmIt can indicate are as follows:
In formula:Represent convolution algorithm；ΔWGmWith Δ WLmRespectively represent node m generator power and load power with
Machine variable；
Step 7.2: according to the homogeneity and additive property of Cumulants method, the algebra that formula (20) is rewritten as Cumulants method being transported
It calculates, the k rank cumulant of node m injecting powerIt can indicate are as follows:
Step 7.3: according to the linearisation tide model of formula (16), obtaining the k rank cumulant Δ X of node voltage and Branch Power Flow(k)With Δ H(k):
Step 7.4: the probability for solving the cumulant of stochastic variable respectively with the method for Gram-Clarlier series expansion is close
The probability density of the cumulant of the probability density and Branch Power Flow of the cumulant of degree, node voltage:
Cumulative distribution function f (x) can be expressed as the series expansion of the all-order derivative containing α (x), then the all-order derivative exhibition of f (x)
Open type are as follows:
In formula, x is each stochastic variable, and α (x) is the probability density function of the stochastic variable of standardized normal distribution；
H in formular(x) it is Xue Fu-Amire spy's multinomial:
Convolution (23) (24) obtains the cumulative distribution function of stochastic variable are as follows:
G is input variable, node voltage and Branch Power Flow and cumulant in formula.
Step 7.5: according to cumulative distribution function and probability density relationship:
F (x) dx=P (26)
Find out the probability distribution of node voltage, power；It is embodied as the probability of each voltage of a certain node a certain moment, performance number
With all moment voltages of a certain node day, power probability maximum value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910255111.XA CN109802394B (en) | 2019-04-01 | 2019-04-01 | Probability load flow calculation method considering distributed power supply and electric vehicle access |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910255111.XA CN109802394B (en) | 2019-04-01 | 2019-04-01 | Probability load flow calculation method considering distributed power supply and electric vehicle access |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109802394A true CN109802394A (en) | 2019-05-24 |
| CN109802394B CN109802394B (en) | 2022-05-17 |
Family
ID=66563136
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910255111.XA Active CN109802394B (en) | 2019-04-01 | 2019-04-01 | Probability load flow calculation method considering distributed power supply and electric vehicle access |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109802394B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110380419A (en) * | 2019-07-11 | 2019-10-25 | 广西电网有限责任公司电力科学研究院 | It is a kind of to mix random and interval variable uncertain tidal current computing method |
| CN110456223A (en) * | 2019-08-19 | 2019-11-15 | 西南交通大学 | A kind of power distribution network short circuit current measuring method containing distributed generation resource and electric car |
| CN110970900A (en) * | 2019-12-10 | 2020-04-07 | 国电南瑞科技股份有限公司 | Evaluation index calculation method for improving voltage stability during phase modulation operation of photo-thermal unit |
| CN110970901A (en) * | 2019-12-11 | 2020-04-07 | 国电南瑞科技股份有限公司 | Control method and system for adapting to voltage safety of fluctuating power supply and direct current transmission receiving terminal |
| CN111146821A (en) * | 2019-12-31 | 2020-05-12 | 国网浙江省电力有限公司嘉兴供电公司 | DSTATCOM optimal configuration method considering photovoltaic uncertainty |
| CN111612248A (en) * | 2020-05-20 | 2020-09-01 | 云南电网有限责任公司电力科学研究院 | Power distribution network side source-load coordination method and system |
| CN111668845A (en) * | 2020-06-16 | 2020-09-15 | 广东工业大学 | Probability load flow calculation method considering photovoltaic correlation |
| CN112636356A (en) * | 2020-11-26 | 2021-04-09 | 广东电网有限责任公司广州供电局 | Method and device for analyzing voltage probability of boundary nodes of main network and distribution network |
| CN112733348A (en) * | 2020-12-30 | 2021-04-30 | 广东电网有限责任公司 | Hybrid power grid probability calculation method and device based on polynomial and maximum entropy model |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106451455A (en) * | 2016-08-29 | 2017-02-22 | 甘肃省电力公司风电技术中心 | Stochastic load flow method containing distributed type power supply system based on node voltage correlation |
| CN106786595A (en) * | 2016-11-29 | 2017-05-31 | 国电南瑞科技股份有限公司 | One kind considers the probabilistic probability load flow calculation method of static frequency characteristic |
| CN107681685A (en) * | 2017-09-13 | 2018-02-09 | 国网甘肃省电力公司电力科学研究院 | A kind of Probabilistic Load computational methods for considering photovoltaic non-linear dependencies |
| CN108493942A (en) * | 2018-04-16 | 2018-09-04 | 华中科技大学 | It is a kind of meter and electric vehicle Probabilistic Load Flow acquisition methods |
| CN109066688A (en) * | 2018-09-06 | 2018-12-21 | 国网安徽省电力有限公司芜湖供电公司 | Based on the Probabilistic Load Flow data capture method under renewable energy uncertainty |
| US20180375332A1 (en) * | 2016-11-24 | 2018-12-27 | China Electric Power Research Institute Company Limited | Method and apparatus for determining distributed power supply access capacity, and storage medium |
| CN109242199A (en) * | 2018-09-27 | 2019-01-18 | 东南大学 | A kind of active load sacurity dispatching method that can be used under the lotus mutual environment of source |
-
2019
- 2019-04-01 CN CN201910255111.XA patent/CN109802394B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106451455A (en) * | 2016-08-29 | 2017-02-22 | 甘肃省电力公司风电技术中心 | Stochastic load flow method containing distributed type power supply system based on node voltage correlation |
| US20180375332A1 (en) * | 2016-11-24 | 2018-12-27 | China Electric Power Research Institute Company Limited | Method and apparatus for determining distributed power supply access capacity, and storage medium |
| CN106786595A (en) * | 2016-11-29 | 2017-05-31 | 国电南瑞科技股份有限公司 | One kind considers the probabilistic probability load flow calculation method of static frequency characteristic |
| CN107681685A (en) * | 2017-09-13 | 2018-02-09 | 国网甘肃省电力公司电力科学研究院 | A kind of Probabilistic Load computational methods for considering photovoltaic non-linear dependencies |
| CN108493942A (en) * | 2018-04-16 | 2018-09-04 | 华中科技大学 | It is a kind of meter and electric vehicle Probabilistic Load Flow acquisition methods |
| CN109066688A (en) * | 2018-09-06 | 2018-12-21 | 国网安徽省电力有限公司芜湖供电公司 | Based on the Probabilistic Load Flow data capture method under renewable energy uncertainty |
| CN109242199A (en) * | 2018-09-27 | 2019-01-18 | 东南大学 | A kind of active load sacurity dispatching method that can be used under the lotus mutual environment of source |
Non-Patent Citations (3)
| Title |
|---|
| 卫鹏等: "基于半不变量和Gram-Charlier级数展开法的随机潮流算法", 《电力工程技术》 * |
| 隋冰彦等: "基于最大熵原理的含风电和电动汽车电力系统概率潮流", 《电网技术》 * |
| 黄煜等: "基于拉丁超立方采样技术的半不变量法随机潮流计算", 《电力自动化设备》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110380419A (en) * | 2019-07-11 | 2019-10-25 | 广西电网有限责任公司电力科学研究院 | It is a kind of to mix random and interval variable uncertain tidal current computing method |
| CN110456223A (en) * | 2019-08-19 | 2019-11-15 | 西南交通大学 | A kind of power distribution network short circuit current measuring method containing distributed generation resource and electric car |
| CN110456223B (en) * | 2019-08-19 | 2020-09-01 | 西南交通大学 | Method for measuring and calculating short-circuit current of power distribution network containing distributed power supply and electric automobile |
| CN110970900B (en) * | 2019-12-10 | 2022-09-20 | 国电南瑞科技股份有限公司 | Evaluation index calculation method for improving voltage stability during phase modulation operation of photo-thermal unit |
| CN110970900A (en) * | 2019-12-10 | 2020-04-07 | 国电南瑞科技股份有限公司 | Evaluation index calculation method for improving voltage stability during phase modulation operation of photo-thermal unit |
| CN110970901A (en) * | 2019-12-11 | 2020-04-07 | 国电南瑞科技股份有限公司 | Control method and system for adapting to voltage safety of fluctuating power supply and direct current transmission receiving terminal |
| CN111146821A (en) * | 2019-12-31 | 2020-05-12 | 国网浙江省电力有限公司嘉兴供电公司 | DSTATCOM optimal configuration method considering photovoltaic uncertainty |
| CN111612248A (en) * | 2020-05-20 | 2020-09-01 | 云南电网有限责任公司电力科学研究院 | Power distribution network side source-load coordination method and system |
| CN111612248B (en) * | 2020-05-20 | 2023-09-08 | 云南电网有限责任公司电力科学研究院 | Power distribution network side source-load coordination method and system |
| CN111668845A (en) * | 2020-06-16 | 2020-09-15 | 广东工业大学 | Probability load flow calculation method considering photovoltaic correlation |
| CN111668845B (en) * | 2020-06-16 | 2023-04-18 | 广东工业大学 | Probability load flow calculation method considering photovoltaic correlation |
| CN112636356A (en) * | 2020-11-26 | 2021-04-09 | 广东电网有限责任公司广州供电局 | Method and device for analyzing voltage probability of boundary nodes of main network and distribution network |
| CN112733348A (en) * | 2020-12-30 | 2021-04-30 | 广东电网有限责任公司 | Hybrid power grid probability calculation method and device based on polynomial and maximum entropy model |
| CN112733348B (en) * | 2020-12-30 | 2022-10-11 | 广东电网有限责任公司 | Hybrid power grid probability calculation method and device based on polynomial and maximum entropy model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109802394B (en) | 2022-05-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109802394A (en) | A kind of probability load flow calculation method counted and distributed generation resource and electric car access | |
| Yan et al. | Stochastic multi-scenario optimization for a hybrid combined cooling, heating and power system considering multi-criteria | |
| Ge et al. | A hybrid model for short-term PV output forecasting based on PCA-GWO-GRNN | |
| Ma et al. | Solar photovoltaic system modeling and performance prediction | |
| CN103353979B (en) | The Optimizing Site Selection of a kind of distributed power source and constant volume method | |
| CN105354655B (en) | Photovoltaic power station group confidence capacity evaluation method considering power correlation | |
| CN103218757B (en) | A kind of method determining photovoltaic generation volume metering | |
| CN107480825B (en) | Photovoltaic power station optimization planning method considering capacity credibility | |
| CN108306303A (en) | A kind of consideration load growth and new energy are contributed random voltage stability assessment method | |
| CN104751006B (en) | It is a kind of meter and correlation of variables probability load flow calculation method | |
| CN107145707A (en) | It is a kind of to count and photovoltaic is exerted oneself the power distribution network transformer planing method of uncertain and overall life cycle cost | |
| Zeng et al. | Forecasting China's hydropower generation capacity using a novel grey combination optimization model | |
| Chang et al. | Interpretable machine learning in sustainable edge computing: A case study of short-term photovoltaic power output prediction | |
| Wu et al. | Reliability evaluation of the solar power system based on the Markov chain method | |
| Huang et al. | An efficient probabilistic approach based on area grey incidence decision making for optimal distributed generation planning | |
| Xiong et al. | Optimal Identification of Unknown Parameters of Photovoltaic Models Using Dual‐Population Gaining‐Sharing Knowledge‐Based Algorithm | |
| CN105633950B (en) | A kind of probabilistic multiple target Random-fuzzy Dynamic Optimal Power Flow Problem method for solving of consideration wind-powered electricity generation injection | |
| Li et al. | Optimal power flow calculation considering large-scale photovoltaic generation correlation | |
| Jiang et al. | Improved gradient‐based optimizer for parameters extraction of photovoltaic models | |
| Al-Dhaifallah | Analytical solutions using special trans functions theory for current–voltage expressions of perovskite solar cells and their approximate equivalent circuits | |
| Ahmed et al. | Non-iterative MPPT Method: A Comparative Study | |
| Thai et al. | Design a photovoltaic simulator system based on two-diode model with linear interpolation method | |
| Yang et al. | Online Network Partition for Voltage Control in Power Systems with Large Penetration of Renewable Energy | |
| Tay et al. | PV parameter identification using reduced IV data | |
| Echarradi et al. | Ecological impact due to the implementation of a modeled and optimized hybrid system. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |